Interstitial therapy with targeted protein toxins for malignant brain tumors: We are investigating a new approach for the treatment of brain tumors which utilizes a new delivery approach for distribution of a class of potent, targeted anti-cancer compounds, called targeted protein toxins. Preclinical in vitro and in vivo experiments of toxins targeted to the transferrin receptor and epidermal growth factor (EGF) demonstrated significant antitumor activity against a variety of tumor types, including malignant gliomas. New methods of drug delivery have been developed to deliver these agents to brain tumors, and in vivo imaging methods are being developed to demonstrate drug distribution in patients. Initially we completed a phase I trial of regional therapy with the targeted protein toxin transferrin-CRM 107 (Tf-CRM107) for the treatment of recurrent malignant brain tumors. Tf-CRM107 is a conjugate of human transferrin (Tf) and diphtheria toxin with a point mutation (CRM107). Tf-CRM107 binds to the Tf receptor, which facilitates iron uptake and is present in higher number on tumor cells than on the normal cells of the brain; the diphtheria toxin mutant kills tumor cells to which the Tf-CRM107 binds. The purpose of the Phase I study was to evaluate the toxicity of Tf-CRM107 when delivered by intratumoral and peritumoral slow interstitial infusion in a dose escalation schedule and to assess antitumor activity in these patients. Twenty-seven patients with malignant brain tumors refractory to standard therapy (surgery, radiation chemotherapy) were treated. The results indicated that therapy with Tf-CRM107 produces tumor responses without severe neurologic or systemic toxicity. A multicenter Phase II study has now finished patient entry. The preliminary results suggest similar efficacy and toxicity as was observed in the Phase I trial. A new protocol to investigate the safety of perfusing the region of brain surrounding tumor excision to eliminate infiltrating tumor cells and to examine the potential of protecting the brain, but not tumor, with this approach combined with systemic administration of chloroquine, an agent that inhibits endothelial cell injury by Tf-CRM 107, is being prepared.Vascular permeability factor/vascular endothelial growth factor in the CNS: Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VPF), is a protein secreted by many cell types which has been shown in model systems unrelated to the central nervous system (CNS) to perform two major functions: it is an angiogenic, endothelial-specific growth factor, and a potent inducer of vascular leakage. Our continuing goals are to define the function of VEGF in the central nervous system, to understand the extent to which altered expression of VEGF contributes to the development of certain pathologic conditions, and to develop strategies for modifying the activity of VEGF in appropriate clinical settings. Recently our work has emphasized investigation of the direct effects of VEGF on normal brain. Since increased VEGF expression has been associated with several CNS pathologies (including brain tumors) these findings have ramifications for the understanding/treatment of the brain edema and inappropriate angiogenesis associated with various CNS diseases. The results indicate:1) Chronic overexposure of normal brain to VEGF greatly increases the capillary permeability to both large and small molecules, resulting in focal blood-brain barrier (BBB) disruption without cell destruction, inflammation, or neurologic deficit. VEGF administration may permit a prolonged, nontoxic method for opening the BBB in order to enhance delivery of a variety of substances into the brain for experimental or therapeutic purposes. Alternatively, specific inhibition VEGF expression/action might restore normal cerebral vascular function in situations where BBB disruption exists. 2) VEGF alone does not elicit an inflammatory response. However, VEGF can liberate an immune response to a foreign antigen (e.g. a virus) already present in the CNS. VEGF potentiates inflammation in CNS not only by opening the BBB and allowing contact between normally sequestered antigens and blood-borne immune mediators, but also by inducing immune mediators such as intercellular adhesion molecules and major histocompatibility complex class I and II expression in the CNS. The modulation of inflammation in the CNS by VEGF may have ramifications for inflammatory diseases of the CNS such as multiple sclerosis. 3) Although VEGF administration is sufficient to induce angiogenesis outside of the nervous system, the normal brain is relatively refractory to VEGF-induced angiogenesis. Other perturbations (e.g. a brain tumor) must be present in conjunction with VEGF in order to induce CNS angiogenesis. This may be explained by differential expression of VEGF receptors in normal brain vessels compared to tumor-associated vessels. The altered profile of VEGF receptors on brain tumor blood vessels allows a selective targeting of tumor vessels through VEGF receptor-specific toxins. However, such toxins are unable to target small infiltrative tumors not yet dependent on VEGF-mediated angiogenesis. Approaches to target tumor blood vessels at earlier stages of development are being sought. Glioblastoma multiforme after whole brain irradiation: A reproducible large animal model of malignant gliomas without the disadvantages imposed by implantation of carcinogens or heterotransplants would enhance investigation of the biology and treatment of infiltrative, intrinsic glial tumors. Radiation has been implicated in the development of primary CNS tumors in humans and animal models. We describe the development of glioblastoma multiforme (GBM) in 7 of 11 rhesus monkeys 0.5-7 years after 35 Gy whole brain radiation. 11 monkeys were treated with fractionated whole brain radiation (350 cGy/day x 10 d). We obtained non- and contrast-enhanced MRIs before treatment and serially after treatment. Animals were sacrificed when they developed neurological symptoms or parenchymal lesions on MRI. 4 animals are alive, three of whom are without MRI abnormalities. The other is currently developing a large lesion in the left fronto-parietal white matter and corpus callosum. 1 animal developed neurological symptoms 2.5 yrs after radiation but had no evidence of radiographic or microscopic disease. All animals that received radiation only developed GBMs, while 2/5 animals treated with pentobarbital and radiation have developed GBMs. 7 animals developed tumors which appeared as contrast-enhancing masses on MRI after 6 months (N=1), 2 y (N=2), and 4 y (N=2), 6.5 y (N=1) and 7y (N=1). 3 animals survived for 0.5, 2 and 3 yrs after development of MRI abnormalities. The lesions were supratentorial (N=5), infratentorial (N=2) and in both compartments (N=1). 3 animals had diffuse bilateral lesions. Six animals have histologically confirmed GBMs; 1 animal with a tumor is being observed with serial MRI. 4 animals had multifocal tumors on histopathological analysis. All tumors demonstrated hypercellularity with giant cells, nuclear pleomorphism, mitoses, neovascularity, necrosis, and pseudopalisading. Extensive infiltration occurred along white matter tracts and along Virchow-Robin spaces. Work is in progress to investigate the pattern of expression of VEGF, transferrin receptor, and other markers of malignant gliomas. This infiltrative glioma model, which has histopathological characteristics identical to gliomas in humans, will be useful for investigation of the anatomy and mechanism of tumor infiltration, for genetic analysis of tumor progression, for detecting unique gene expression by infiltrating tumor cells, for determining unique cell-surface receptors and biochemistry that may permit selective elimination of infiltrating tumor cells, and for examining new treatments for infiltrative glial tumors Studies of human pituitary tumors: We continue to investigate new approaches to diagnose Cushing syndrome (CS), establish the differential diagnosis of CS, and localize small pituitary adenomas to aid in the diagnosis and treatment of patients with CS. One of the reasons for failure of surgery in Cushing&#8217;s disease(CD) is that many patients have small tumors, tumors that are difficult to detect with imaging techniques and that are difficult to find at surgery. Only 55% of patients with surgically-proven microadenomas have positive MRI. Intraoperative ultrasound, using a prototype 12 Mhz probe developed for transsphenoidal surgery, was shown to detect and localizing very small tumors in the pituitary gland during surgery. Adenomas were detected in 70% of patients with normal sellar MRI scans. The use of these techniques resulted in curative pituitary surgery in 95% of a large group of patiens with Cushings disease and negative MRI, the subset of patients that has been difficult to treat successfully. Several years ago we developed and examined the use of inferior petrosal sinus (IPS) sampling to diagnose CD. We recently demonstrated that 1) with the use of corticotropin releasing hormone (CRH), jugular venous sampling, an easier procedure to perform, provides the correct diagnosis in many patients, potentially eliminating the need for IPS sampling, and 2) a hypoplastic IPS, which can be detected during the procedure, is the explanation for the very rare error in diagnosis with bilateral IPS sampling.Ectopic ACTH-secreting pituitary tumors that originated in the cavernous sinus were identified and cured in 5 patients with CD. This observation suggests that patients in whom no tumor can be found at surgery, even with total hypophysectomy, and who respond to sellar irradiation may have an ectopic parasellar pituitary tumor in the field of irradiation as the explanation of the remission of CS.Thyroid-stimulating hormone (TSH) &#8211; secreting pituitary tumors are the most aggressive type of pituitary tumor and the most difficult to treat successfully. One reason for this is a late diagnosis and the invasive nature of these tumors. Techniques for diagnosis and treatment of these tumors were compared and treatment at an earlier stage was shown to enhance the likelihood of curative therapy.Molecular techniques were used to demonstrate allelic loss of 11q13 in benign pituitary tumors of patients with the Multiple Endocrine Neoplasia I (MEN I) syndrome. Furthermore, 2 tumors were shown to originate from pluripotent tumor cells in the pituitary, potentially explaining the basis of the origin of multiple types of pituitary tumors in these patients. MEN I gene alterations were detected in a small subset of pituitary tumors in patients without MENI, indicating that some, but not most, sporadic pituitary tumors arise as a result of a mutation or deletion of this gene.